Development of forming condition optimization methodology in the deep drawing process for automotive outer parts

Abstract

Sheet metal forming simulation is now widely used for predicting forming defects in the design of drawing die for automotive panel. For defect countermeasure, however, performing a variety of simulation leads to requiring many man-hours with trial and error process. The numerical optimization methodology has been recently applied to stamping as research. Optimization of forming conditions such as the actual parts' draw-bead force, blank size, etc. results in difficulty, because it involves a lot of constraints such as formability and cost. Moreover, it has multiple peaks. This research focused on variables independence and developed the new method to search the best feasible solution in case which has the complicated design problems such as an actual automotive panel. There are two steps in the developed method: 1) Optimization process is applied to the design variables with strongly related objective functions; 2) All design variables are optimized by considering all objective functions. This method was compared with such optimization techniques as Simulated Annealing, etc. The optimized solutions for automotive parts can be obtained only by applying the proposed method.